- From: Benjamin C. Wiley Sittler <notifications@github.com>
- Date: Wed, 10 Jun 2015 20:47:02 -0700
- To: slightlyoff/ServiceWorker <ServiceWorker@noreply.github.com>
- Message-ID: <slightlyoff/ServiceWorker/issues/707/110991693@github.com>
Given that, async map with full access to exactly the same data as document.cookie with no new functionality (e.g. it would reveal on;y names and values you could see in document.cookie, and setting would allow you the same domain, path, name, value, maxage, secure, httponly and even expires control [the cookie expiration may in fact need to be a precise date and time!) and no new functionality (e.g. reading interface only exposes in-scope cookie names and values in the same order as document.cookie and does not tell you their domain, path, expiration time, or secure flag) and just a cleaner API (no string parsing, no concatenation, allow a date object for expiration) would seem to me like the obvious starting point. On Jun 10, 2015 15:52, "Alex Russell" <notifications@github.com> wrote: > A few points: > > - having access to cookies is a no-brainer. We can't prevent it (if > there's a document, there's an API that already does it already); can we > just get over our bad selves and get on with a design that does cookies > better? > - extra restrictions over what the current API can do need to be > justified; the burden of proof is on removal, not equality of functionality > - setting should be symmetric; if reading is async, setting should be > too > > — > Reply to this email directly or view it on GitHub > <https://github.com/slightlyoff/ServiceWorker/issues/707#issuecomment-110940363> > . > --- Reply to this email directly or view it on GitHub: https://github.com/slightlyoff/ServiceWorker/issues/707#issuecomment-110991693
Received on Thursday, 11 June 2015 03:47:33 UTC